Skip to main navigation menu Skip to main content Skip to site footer

Vol. 11 No. 1 (2025)
Special Issue: Proceedings of CVIS 2025

Articles

Whole-Slide Image Compression and On-Demand Viewing using Reference-Based Super-Resolution

https://doi.org/10.15353/jcvis.v11i1.10017

Published 2026-04-01

  • Wenbo Yang
    • ,
  • Seungho Shin
    • ,
  • Richard Y. Zhu
    • ,
  • Zhou Wang

Abstract

The transition from glass slide to digital pathology has been hindered by the vast storage and transmission requirements of archiving and remote-accessing whole-slide images (WSIs): a small regional hospital may generate hundreds to thousands of gigabytes of WSIs per day. Neural compression methods show good promise but suffers from potential hallucinations. Here we examine and benchmark modern statistical compression methods and develop a WSI archival pipeline that incorporates both traditional and machine learning techniques, achieving 93\% reduction in space compared to conventional JPEG compression. The proposed pipeline is designed to be fallback-safe: a computer without machine-learning capabilities can still view archived WSIs without content alteration, while substantially more details at target regions on-demand may be recovered by our real-time reference-based super-resolution upsampling method.
PDF